GEO (Generative Engine Optimization): Content Strategy for Claude and ChatGPT

Ranking on Google's first page is no longer the only goal. In 2024, 43% of users ask ChatGPT, Claude, or Perplexity first when they want to learn something. Do these engines cite sources when answering? If your site isn't in these citations, you're losing traffic. At FUTIA, I've built content infrastructure for over 40 Turkish brands in the last six months, and here's what I've observed: sites doing classic SEO don't appear in AI engines, but those investing in structured data and content depth get cited. In this post, I'll explain what GEO is, why it works differently from SEO, and what steps you need to take for ChatGPT or Claude to use your site as a source. I'll also share a real case study from FUTIA, showing how we structured 618 recipe contents and opened them up to AI citations.

What is GEO (Generative Engine Optimization)?

GEO is the practice of optimizing your content for large language models (LLMs). In classic SEO, Google's bots crawl your site, index the content, and rank it according to search queries. In GEO, AI engines like ChatGPT, Claude, and Perplexity use your content as a reference source and cite it when answering users. The difference: Google lists you, AI directly includes you in the answer.

An example: when you ask ChatGPT "How do you make Turkish stuffed grape leaves with olive oil?" the model tells you the recipe and adds 2-3 sites as sources below. If your site isn't one of those sources, that user never becomes aware of your existence. This is exactly where GEO comes into play.

The fundamental difference of GEO is that it focuses on machine readability rather than human readability. Schema.org markup, JSON-LD structures, clear listing formats, clear headings, and source references form the backbone of GEO. While "keyword density" matters in SEO, "data structure" and "contextual clarity" matter in GEO.

At FUTIA, when I produced 618 recipe contents for italyanmutfagi.com, I added Schema.org Recipe markup to each one. Result: Claude and ChatGPT started showing these recipes as sources. Because models can parse structured data more easily and find it reliable.

Why aren't ChatGPT and Claude citing your site?

Most sites aren't cited by AI engines because content isn't structured, semantic context is weak, and source reliability is unclear. Three main reasons stand out:

1. Lack of structured data

When ChatGPT or Claude scans a page, it primarily processes structured data, not plain text in HTML. If your site lacks Schema.org markup, the model evaluates your content as "difficult to understand." For example, you wrote a recipe post, but ingredients aren't in list format—they're mentioned within paragraphs. The model can't parse this and won't use it as a source.

At doktorbul.com, I added Person and MedicalBusiness schema to each of 79,000 doctor profiles. Result: for queries like "a good cardiologist in Istanbul," Claude started showing doktorbul.com as a source.

2. Insufficient content depth

AI engines don't cite superficial content. A 300-word blog post has no reference value for ChatGPT. The model prefers comprehensive content divided into subheadings and supported with examples. Also, content recency matters. An article written in 2019 has a low chance of being cited in 2025.

At FUTIA, we expanded product pages for diolivo.com.tr. We added "how to use," "suitable for whom," and "frequently asked questions" sections to each product. In six months, traffic increased 340% and ChatGPT started showing diolivo.com.tr as a source for olive oil products.

3. Unclear source reliability

AI engines look at backlink profile, domain age, HTTPS status, and content consistency to understand how reliable your site is. If your site is new, has low backlink count, or content quality fluctuates, the model avoids using you as a source.

At memuratamalari.com, we achieved 40,400 monthly organic search traffic because we published current job listing content regularly every day. This process, which we automated with Claude Haiku API, increased the site's reliability score and it started appearing in AI citations.

How to build a content strategy for GEO?

Success in GEO comes from writing content not just for humans, but for both humans and machines. I recommend following these steps:

Add structured data

Use appropriate Schema.org markup for each content type. If you're writing a recipe, add Recipe schema; if you're creating a product page, add Product schema; if you're writing an article, add Article schema. JSON-LD format is preferred because both Google and LLMs process it more easily.

Example: for italyanmutfagi.com, we used this structure for each recipe:

• recipeIngredient: list of ingredients
• recipeInstructions: step-by-step instructions
• totalTime: preparation and cooking time
• recipeYield: serves how many
• nutrition: nutritional values

Thanks to this structure, when Claude answered "How do you make Tiramisu?" it showed italyanmutfagi.com as a source.

Deepen content

Each page should be at least 1,200 words. Use subheadings (H2, H3), prefer list formats, give examples. Also add a "frequently asked questions" section because AI engines can directly cite this section.

At kamupersonelhaber.com, I added sections like "Who is this listing suitable for?", "What are the application requirements?", "How does the interview process work?" to each job listing page. Result: ChatGPT started showing kamupersonelhaber.com as a source to those asking about public job listings.

Maintain freshness

AI engines don't cite old content. Update your content regularly. Add timestamps (using lastModified schema) and keep adding new information.

At FUTIA, we produced 2,000+ video contents for futia.net in three months and continued adding new content every week. This consistency increased the site's visibility in AI engines.

Use source references

Link to other reliable sources in your own content. AI engines find sites that provide sources more reliable. Also, outbound links show that your site is an "information hub."

Optimize technical infrastructure

Site speed, mobile compatibility, HTTPS, XML sitemap, robots.txt should be in order. AI bots struggle to crawl technically problematic sites and won't cite them.

Differences between GEO and SEO

Although GEO and SEO seem similar, their target audiences and metrics are different. In SEO, the target is Google's algorithm; in GEO, the target is LLMs' training data and real-time crawling processes.

In SEO, you do keyword research; in GEO, you think "what questions are we answering?" In SEO, backlink count matters; in GEO, content structure and depth matter. In SEO, you track rankings; in GEO, you check "which AI engines are citing us?"

Another difference: in SEO, users click to your site; in GEO, users get the answer from AI and see you as a source. So in GEO, traffic increase may not be direct, but brand awareness and authority increase.

At FUTIA, I invest in both strategies. For diolivo.com.tr, we both improved Google rankings (SEO) and got into ChatGPT citations (GEO). Result: organic traffic increased 340% and brand awareness multiplied.

What tools should you use for GEO?

Since GEO is still a new field, specialized tools are limited. But you can measure your GEO performance using these tools:

• Schema Markup Validator: Google's tool checks your structured data.
• ChatGPT and Claude: Query your own site, check if it's shown as a source.
• Perplexity: Search your topic titles on Perplexity, see if your site is in the sources.
• Google Search Console: Track which queries you appear in, optimize similar queries for GEO.
• Screaming Frog: Detect technical SEO issues, remove crawling barriers for AI bots.

At FUTIA, I prepare monthly GEO reports for clients. I do manual checks on ChatGPT, Claude, and Perplexity, listing which pages we're cited on. I also detect and fix Schema.org deficiencies.

Real case study: GEO implementation at italyanmutfagi.com

We produced 618 recipe contents for italyanmutfagi.com. We used Claude Sonnet 3.5 for each recipe, but took these steps to structure the content:

1. Schema.org Recipe markup: We added Recipe schema in JSON-LD format to each recipe. Ingredients, steps, time, nutritional values were structurally marked. 2. Subheadings and lists: Each recipe was divided into "Ingredients," "Preparation," "Tips," "Frequently Asked Questions" sections. 3. Image optimization: High-quality images were added to each recipe, alt texts were written. 4. Internal linking: A "Similar Recipes" section was added between recipes, making navigation easier for users and bots. 5. Keeping current: New recipes were added regularly, existing recipes were updated.

Result: When you ask ChatGPT about Italian cuisine, italyanmutfagi.com is shown as a source. Claude cites it similarly. Site traffic increased 180% in the first three months and also rose in Google rankings.

This case taught me this: GEO doesn't replace SEO, it complements it. Structured data is beneficial for both AI engines and Google. Content depth increases both user experience and citation chances.

Common mistakes in GEO

There are some mistakes that GEO beginners make. I recommend avoiding these:

Using structured data incorrectly

Adding Schema.org markup isn't enough, you need to use it correctly. For example, if you add Article schema to a recipe post, AI engines won't recognize it as a recipe. Use appropriate schema for each content type.

Writing content only for AI

GEO doesn't mean writing content for machines instead of humans. Content should be both readable and structured. Overly technical or robot-like written content ruins user experience and lowers conversion rates.

Neglecting freshness

Doing GEO optimization once and leaving it isn't enough. You should update content regularly, add new information, update dates. AI engines don't cite old content.

Focusing on only one AI engine

Content you optimize for ChatGPT may not work on Claude. Each engine has different preferences. Generally, structured data, content depth, and source reliability work on every engine.

The future of GEO: 2025 and beyond

GEO is still in its infancy, but growing rapidly. In 2025, AI engine usage will increase, Google's market share will decrease. This will further increase GEO's importance.

At FUTIA, I'm tracking these trends:

• Multimodal content: AI engines are starting to process not just text, but images, videos, and audio. Make your content multimodal.
• Real-time data: AI engines prefer real-time data over static content. API integrations and live data streams will gain importance.
• Personalized answers: AI engines give user-specific answers. Optimize your content for different user segments.
• Voice search: ChatGPT and Claude are developing voice search features. Optimize your content for voice queries.

To adapt to these trends, you need to act now. GEO will be one of the most important strategies providing competitive advantage in 2025.

How do you prepare your site for GEO?

GEO optimization requires technical knowledge and continuous updates. If doing this on your own seems difficult, we at FUTIA can help. I'm Miraç, and in the last six months I've built GEO infrastructure for over 40 Turkish brands. On projects like doktorbul.com, diolivo.com.tr, and italyanmutfagi.com, I've implemented structured data, content depth, and AI citation strategies.

At FUTIA, we do everything from site setup to automation, from monthly maintenance to content production. Claude and ChatGPT API integrations, Schema.org markup, programmatic SEO, cart abandonment automations are our areas of expertise. I work from the Netherlands but provide full support to Turkish brands.

If you want to evaluate your site's GEO potential, you can contact me via WhatsApp: +90 532 491 17 05. Or send an email: info@futia.net. In the first meeting, I'll analyze your site and explain in detail what steps you need to take. GEO is the new way to be visible in 2025, and those who start early on this path will win.